graph
Toolbox module for working with networkx graphs.
Module contains functions for calculating graph centrality, visualizing
graphs and finding various network properties, in addition to various
other useful functions.
Graph centralities are accessed using the centralities() function, which
takes as arguments a graph and the metric to use as a constant of the
GraphMetrics class.
-
class graph.GraphMetrics
- Class holding constants for the different graph centrality metrics
-
graph.add_edges_from_matrix(graph, matrix, nodes, rel_weight=1.0)
Add edges to graph based on adjacency matrix.
The nodes list corresponds to each row/column in matrix.
The rel_weight scales the edge weights.
-
graph.betweenness(G)
- Betweenness centrality
-
graph.centralities(graph, method)
- Return centralities for nodes in graph using the given centrality method
-
graph.closeness(G)
- Closeness centrality
-
graph.clustering_degree(G)
Clustering degree ‘’centrality’‘
Measure of centrality based on the clustering coefficient of a node
multiplied with its weighted degree centrality.
-
graph.current_flow_betweenness(G)
- Current-flow betweenness centrality
-
graph.current_flow_closeness(G)
- Current-flow closeness centrality
-
graph.draw(graph)
- Draw the graph
-
graph.draw_with_centrality(G, sizeV=None, min_size=500, max_size=4000, default_size=500, layout=None)
- Visualizes a graph preserving edge length and demonstrating centralities as node sizes.
-
graph.equal(g1, g2)
Check if two graphs are identical.
The graphs are considered equal if they contain the same set of nodes
and edges. Edge weights are not considered.
-
graph.get_hubs(graph, n=10)
- Return the n most important hubs from the graph
-
graph.hits_authorities(G)
- The HITS authorities centralities
-
graph.hits_hubs(G)
- The HITS hubs centralities
-
graph.invert_edge_weights(G)
- Returns a graph with all edge weights inverted
-
graph.load(G)
- Load centrality
-
graph.network_properties(graph, plot_distribution=False, verbose=False)
- Returns information about the graph.
-
graph.node_set(graphs)
- Return list of unique nodes from list of graphs
-
graph.normalize(A)
- Normalize a numpy nd-array
- PageRank values for nodes in graph G
-
graph.reduce_edge_set(graph, remove_label)
Return new graph with some edges removed.
Those edges that have the remove_label, and no other labels, are removed.
-
graph.weighted_betweenness(G)
- Weighted version of the betweenness centrality
-
graph.weighted_closeness(G)
- Weighted version of the closeness centrality
-
graph.weighted_degree(G, normalize=True)
Weighted degree centralities.
Counts both incomming and outgoing links.
This is the same as the sum of the weighted in-degree and weighted out-degree.
Assumes digraph.
-
graph.weighted_in_degree(G, normalize=True)
- Weighted in-degree centralities
-
graph.weighted_load(G)
- Weighted version of the load centrality
-
graph.weighted_out_degree(G, normalize=True)
- Weighted out-degree centralities